Activation of Wild-Type Cytochrome P450BM3 by the Next Generation of Decoy Molecules: Enhanced Hydroxylation of Gaseous Alkanes and Crystallographic Evidence

Article abstract of DOI:10.1021/cs501592f

The direct hydroxylation of alkanes under mild conditions is a key issue in catalytic chemistry that addresses an increasing number of industrial and economic requirements. Cytochrome P450s are monooxygenases that are capable of oxidizing less reactive C-H bonds; however, wild-type P450s are unavailable for many important nonnative substrates such as gaseous alkanes. Here, we report the enhanced hydroxylation activities and crystallographic evidence for the role of decoy molecules in wild-type P450BM3-catalyzed hydroxylation of gaseous ethane and propane by using the next generation of decoy molecule. A cocrystal structure of P450BM3 and a decoy molecule reveals that an N-perfluoroacyl amino acid (decoy molecule) partially occupies the substrate-binding site of P450BM3. This binding of the decoy re-forms the active site pocket to allow the accommodation of small substrates and simultaneously influences the formation of compound I species by expelling water molecules from the active site. (Chemical Equation Presented).

Full text of DOI:10.1021/cs501592f

Research Article
pubs.acs.org/acscatalysis
Activation of Wild-Type Cytochrome P450BM3 by the Next
Generation of Decoy Molecules: Enhanced Hydroxylation of Gaseous
Alkanes and Crystallographic Evidence
Zhiqi Cong,† Osami Shoji,*,† Chie Kasai,† Norifumi Kawakami,‡ Hiroshi Sugimoto, § Yoshitsugu Shiro, §
and Yoshihito Watanabe*,‡
‡
†Department of Chemistry, Graduate School of Science, and Research Center for Materials Science, Nagoya University, Furo-cho,
Chikusa-ku, Nagoya 464-8602, Japan
§RIKEN SPing-8 Center, Harima Institute. 1-1-1 Kouto, Sayo, Hyogo 679-5148, Japan
S
* Supporting Information
ABSTRACT: The direct hydroxylation of alkanes under mild conditions is a key
issue in catalytic chemistry that addresses an increasing number of industrial and
economic requirements. Cytochrome P450s are monooxygenases that are capable
of oxidizing less reactive C−H bonds; however, wild-type P450s are unavailable
for many important nonnative substrates such as gaseous alkanes. Here, we report
the enhanced hydroxylation activities and crystallographic evidence for the role of
decoy molecules in wild-type P450BM3-catalyzed hydroxylation of gaseous
ethane and propane by using the next generation of decoy molecule. A cocrystal
structure of P450BM3 and a decoy molecule reveals that an N-perfluoroacyl
amino acid (decoy molecule) partially occupies the substrate-binding site of
P450BM3. This binding of the decoy re-forms the active site pocket to allow the
accommodation of small substrates and simultaneously influences the formation
of compound I species by expelling water molecules from the active site.
KEYWORDS: biocatalysis, gaseous alkanes, cytochromes, hydroxylation, decoy molecule
to ensure their specificity and efficiency. For the past few
decades, protein engineering has emerged as a powerful tool to
remodel P450s for the oxyfunctionalization of nonnative
substrates. These studies have succeeded in changing the
substrate specificity and have improved the catalytic activity of
P450s by the application of rational and semirational structure-
based mutagenesis, as well as by directed evolution techniques
(random mutagenesis).10−16 Although challenging, it is still
possible to apply wild-type P450s for the oxidation of small
nonnative substrates.
In the absence of substrates, the generation of an active
intermediate of P450s is prevented by the coordination of a
sixth water ligand in the low-spin resting state of the heme iron
(substrate-free form).17 The binding of substrate perturbs the
aqua ligand, which leads to a shift of the heme to a high-spin
state (substrate-bound form)18 accompanied by a positive
redox potential shift that initiates the reduction of ferric heme
(Figure 1A). Subsequent O2 binding to the ferrous heme
followed by its reductive activation generates an iron(IV)−oxo
cation radical intermediate (compound I), which is generally
considered to be the active species that is responsible for
substrate oxidation (Figure 1B).19,20 According to the oxygen-
INTRODUCTION
■
Gaseous alkanes, which are found in abundance in natural gas,
are important fuels and potential chemical feedstock.1 The
development of molecular catalysts (homogeneous and/or
heterogeneous) for the selective hydroxylation of less-reactive
gaseous alkanes is a longstanding challenge and a current topic
of interest, considering increasing industrial and economic
requirements.2,3 One of the main difficulties for gaseous alkane
hydroxylation is their high C−H bond-dissociation energies
(BDEs). For example, methane and ethane have BDEs of 104.9
and 101.4 kcal/mol, respectively.4 Although methane mono-
oxygenases from methanotrophic bacteria oxidize methane to
methanol under mild conditions, they are not suitable for
further application as biocatalysts, mainly because of difficulties
associated with preparing the multicomponent system in
recombinant form.5,6 Cytochrome P450s (P450s) are a
superfamily of iron-containing monooxygenases that are
capable of breaking strong C−H bonds of hydrocarbons.
Because recombinant forms of P450s are prepared by using
common expression systems such as Escherichia coli, P450s have
great potential for further engineering and subsequent
application in synthetic chemistry.7−9 Unfortunately, the
substrate specificity of cytochrome P450s makes them
unsuitable for the hydroxylation of gaseous small alkanes,
because P450s, especially those isolated from bacteria,
recognize their specific substrates by intermolecular interactions
Received: October 15, 2014
Revised: November 12, 2014
Published: November 18, 2014
© 2014 American Chemical Society
150
dx.doi.org/10.1021/cs501592f | ACS Catal. 2015, 5, 150−156

ACS Catalysis
Research Article
Figure 1. (A) Active site structures of wild-type P450BM3: substrate-free form17 (left, inactive state, PDB code 2HDP); N-palmitoylglycine-bound
form18 (right, active state, PDB code 1JPZ). (B) Catalytic cycle of cytochrome P450. (C) Schematic representation of a plausible catalytic cycle of
the “P450BM3-decoy molecule system”.
activation mechanism of P450s for the generation of compound
I, substrate binding serves as a switch to start the reaction. For
example, P450BM3, a fatty acid hydroxylase isolated from
Bacillus megaterium, selectively binds long-alkyl-chain fatty acids
to start its catalytic cycle.21 Although wild-type P450BM3
shows very high catalytic activity for fatty acid hydroxylation, it
does not efficiently catalyze nonnative substrates with
structures that are significantly different from those of fatty
acids because these nonnative substrates do not bind to the
enzyme to initiate the first step of the catalytic cycle (Figure
1C). The crystal structure of a substrate-bound form of
P450BM3 shows that the fatty acid (palmitoleic acid) is fixed
by two major interactions with P450BM3: (1) a hydrophobic
interaction of the alkyl chain of the fatty acid with hydrophobic
residues in the active site and (2) hydrogen bonding and
electrostatic interaction of the carboxylate in the fatty acid with
amino acids Tyr51 and Arg47 located at the entrance to the
active site of P450BM3 (Figure 2A).22 Insertion of the alkyl
chain into the active site is critical, because this switches on the
reaction by removing the water molecule coordinated to the
ferric heme. In fact, carboxylic acids with a short alkyl chain of
less than 12 carbon atoms are not hydroxylated by P450BM3.23
Furthermore, P450BM3 does not start the catalytic hydrox-
ylation of normal alkanes such as tetradecane and hexadecane,
which do not have the carboxyl group.24 This indicates that the
interaction of the carboxylate with Arg47 and Tyr51 is also
critical for initiating the monooxygenation. This cooperative
switching mechanism of the substrate recognition by P450BM3
contributes to its high substrate specificity. Recently, we have
demonstrated that the switch of P450s can be turned on by
using a series of perfluorinated carboxylic acids (PFCs) with
shorter alkyl chains named “decoy molecules” such as
perfluorinated nonanoic acid.25,26 In these cases, P450BM3
misrecognizes PFCs as its natural substrate because their
structures resemble those of fatty acids, since the atomic radius
of fluorine is similar to that of hydrogen. Thus, PFCs turn on
P450s, which adopt the active, high-spin state to start the
catalytic cycle (Figure 1C). By utilizing this substrate
misrecognition, a variety of substrates such as propane and
benzene can be hydroxylated by wild-type P450BM3.25−28 The
catalytic activities for nonnative substrates, however, are still
lower than those of long-alkyl-chain fatty acids. We presume
the interaction of PFCs with P450BM3 is not sufficient to fully
switch on P450BM3 because of its weak binding affinity to the
enzyme. However, it was reported that hydroxylation of N-
palmitoyl glycine is more efficiently catalyzed by P450BM3
than palmitic acid because of an increased association constant
of N-palmitoyl glycine (Kd = 262 nM).18 The crystal structure
of an N-palmitoylglycine-bound form of P450BM3 shows
additional hydrogen-bonding interactions of the carboxylate
group in N-palmitoylglycine with Gln73 and Ala74 (Figure
2B),18 which are expected to contribute to the improved
binding affinity of N-palmitoylglycine. These observations allow
a new generation of decoy molecules to be developed on the
Figure 2. Substrate-binding sites of P450BM3 in complexes with (A)
palmitoleic acid (PAM, PDB code 1FAG) and (B) N-palmitoylglycine
(NPG, PDB code 1JPZ).
151
dx.doi.org/10.1021/cs501592f | ACS Catal. 2015, 5, 150−156

ACS Catalysis
Research Article
basis of the introduction of an amino acid onto the PFCs
(Chart 1). The resulting molecule was anticipated to interact
strongly with P450BM3, resulting in higher rates of
hydroxylation of small aliphatic and aromatic hydrocarbons.
environment at the entrance of the substrate access channel in
P450BM3 would be conducive for the hydroxylation of
hydrophobic substrates. We thus prepared a series of N-
perfluoroacyl amino acids with hydrophobic side chains as a
new generation of decoy molecules (Chart 1). Herein, we
report that N-perfluoroacyl amino acids with hydrophobic side
chains do indeed strongly activate (arouse) wild-type
P450BM3, and we demonstrate that the hydroxylation rates
of gaseous alkanes are reached to 45/min/P450 for ethane and
256/min/P450 for propane, by employing N-perfluoronona-
noyl-L-leucine. Furthermore, we have succeeded in obtaining
the crystal structure of N-perfluorononanoyl L-tryptophan-
bound P450BM3, which provides crucial evidence for the roles
of decoy molecules and gives some mechanistic insights into
the activation of wild-type P450 by decoy molecules and into
how gaseous alkanes gain access to the heme pocket.
Chart 1. Design of N-Perfluoroacyl Amino Acids as Second-
Generation Decoy Molecules Used in This Study
EXPERIMENTAL SECTION
■
Decoy Molecules. The preparation, spectroscopic charac-
terization, and dissociation constants of decoy molecules, as
well as absorption spectral studies of P450BM3 in the presence
of decoy molecules, are described in the Supporting
Information.
In a mutagenesis study on the oxidation of hydrophobic
substrates by P450BM3,29 the replacement of Arg47 and Tyr51
by hydrophobic amino acids (R47L/Y51F) greatly improved
the catalytic activity. The results indicate that a hydrophobic
Figure 3. Hydroxylation of propane (A), ethane (B), and cyclohexane (C) catalyzed by wild-type P450BM3 (0.5 μM) with N-perfluoroacyl amino
acids (100 μM) as second-generation decoy molecules. The previous results from the first-generation decoy molecules (PFCs) are also shown in
(A). The # symbols denote the reactions in which the product was not detected. N-Perfluorononanoyl amino acids are abbreviated to the three-letter
codes of the corresponding amino acids in (B) and (C). For the hydroxylation of propane (A) and cyclohexane (C), 5 mM NADPH was added to
the reaction mixture. The ethane hydroxylation was performed by using an NADPH regeneration system consisting of D-glucose (100 mM), NADP+
(1 mM), and glucose dehydrogenase (1 unit). The PFR is given in μmol min−1 (μmol of P450)−1, and the error bars denote the standard deviation
calculated from three parallel experiments. The coupling efficiencies ([product formation]/[NADPH consumption] × 100) are shown as red open
circles in (A) and (C).
152
dx.doi.org/10.1021/cs501592f | ACS Catal. 2015, 5, 150−156

ACS Catalysis
Research Article
Procedure for Hydroxylation of Propane. To a stirred
gas-saturated buffer solution (propane/oxygen 80/20 v/v)
composed of 20 mM Tris HCl (pH 7.4) and 100 mM KCl (1
mL) in a glass sample bottle at room temperature were
successively added P450BM3 (0.5 μM), a methanolic solution
of N-perfluoroacyl amino acid (decoy molecule) (100 μM), and
NADPH (5 mM). The reactor was immediately fitted with a
balloon (propane/oxygen ca. 1/1), and air in the bottle was
replaced by the mixed gas for 2 min. After addition of NADPH
to start the reaction, the reaction was carried out for 10 min,
and then a sample of the reaction mixture (10 μL) was taken to
determine the consumption of NADPH. The reaction solution
was mixed with sodium nitrite (150 mg), hexane (1 mL), and
3-pentanol (1 mM, internal standard). Aqueous sulfuric acid
(20%, 150 μL) was added dropwise to the mixture cooled in an
ice bath over 15 min. The organic phase was separated and
washed with water, and the obtained solution was directly
analyzed by gas chromatography (GC2014; Shimadzu) with an
Rtx-1 column (Restek).
a Dundee PRODRG server34 and used in the refinement with
COOT and REFMAC5. All protein figures were depicted by
using PyMOL.35 The final refinement statistics are summarized
in Table S1 (Supporting Information).
RESULTS AND DISCUSSION
■
A series of N-perfluoroacylglycines with different lengths of
alkyl chain (PFC6-Gly to PFC11-Gly) were synthesized both to
examine the effect of the glycine on propane hydroxylation and
to establish an appropriate length for the alkyl chain of the N-
perfluoroacyl amino acids (Figure 3A and pages S1−S4 in the
Supporting Information).36 As expected, the catalytic hydrox-
ylation activity and the binding affinity (Table 1) were both
Table 1. Dissociation Constants of Decoy Molecules Bound
to P450BM3
a
b
second generation
Kd/μM
first generation
Kd/μM
PFC7-Gly
911
95
22
3
PFC8-Gly
PFC8
1900
980
290
91
The hydroxylations of ethane and cyclohexane performed by
using similar procedures are described in detail in the
Supporting Information.
PFC9-Gly
PFC9
PFC10-Gly
PFC11-Gly
PFC9-L-Ala
PFC9-L-Val
PFC9-L-Ile
PFC9-L-Leu
PFC9-L-Met
PFC9-L-Phe
PFC9-L-Trp
PFC10
PFC11
PFC12
PFC13
PFC14
5
Procedure for Hydroxylation of Ethane using an
NADPH Regeneration System. To a stirred gas-saturated
buffer solution (ethane/oxygen 80/20 v/v) composed of 20
mM Tris HCl (pH 7.4) and 100 mM KCl (1 mL) in a glass
sample bottle at room temperature were successively added
P450BM3 (0.5 μM), N-perfluoroacyl amino acid (decoy
molecule) (100 μM), D-glucose (100 mM), NADP+ (1 mM),
and glucose dehydrogenase (1 unit). The reactor was
immediately fitted with a balloon (ethane/oxygen ca. 1/1),
and the reaction was carried out under balloon pressure for 4
min. The reaction solution was mixed with sodium nitrite (150
mg), hexane (1 mL), and 3-pentanol (1 mM, internal
standard), and then the mixture was carefully treated by
addition of diluted sulfuric acid (20%, 150 μL) cooled in an ice
bath. After more than 15 min, the organic phase was separated
and washed with water. The obtained solution was directly
analyzed by gas chromatography (GC2014; Shimadzu) with an
Rtx-1 column (Restek). Control experiments were carried out
with nitrogen gas instead of ethane under the same conditions.
Crystallization of P450BM3 with N-Perfluorononano-
yl-L-Tryptophan. The buffer of the purified P450BM3 was
exchanged with 50 mM Tris-HCl (pH 7.4) containing 100 μM
of N-perfluorononanoyl-L-tryptophan and 1% (v/v) dimethyl
sulfoxide. P450BM3 was then concentrated to 20.3 mg/mL by
centrifugation using Amicon Ultra filter units (Millipore,Co.).
An aliquot of the concentrated P450BM3 solution (1 μM) was
mixed with 1 μM of a reservoir solution composed of 100 mM
Tris-HCl (pH 8.5), 210 mM MgCl, and 21% (w/v) PEG8000.
The cocrystals were grown by a sitting-drop vapor diffusion
method at 20 °C for 20 days.
26
7.3
4.1
39
20
3.4
1.6
30
1.8
13
a
This study. The Kd values of PFC9-DL-Ala and PFC9-D-Ala were not
estimated because of the irregular spectral change during titration
b
experiments. Previous study.25
drastically improved by simply introducing glycine. The
optimum chain length was PFC9, and the use of this compound
gave a product formation rate (PFR) of 2-propanol of 128/
min/P450 (Figure 3A). To examine the effect of the side chain
structure and chirality of the amino acid on propane
hydroxylation, we prepared a series of PFC9 derivatives
modified with amino acids bearing a range of hydrophobic
side chains (Chart 1, Figure 3A, and pages S1−S4 in the
Supporting Information). Among the N-perfluorononanoyl
amino acids examined, N-perfluorononanoyl-L-leucine was the
most effective for propane hydroxylation (PFR = 256/min/
P450, Figure 3A). This activity reaches about half of that
obtained by the best reported mutant of P450BM3 (propane
monooxygenases, P450PMO) with 22 mutations from 15 rounds
of directed evolution (455/min/P450) (Figure S5 and Tables
S2−S5, Supporting Information),37 whereas two PFRs obtained
under different conditions could not be compared directly. This
activity of PFC9-L-Leu for propane hydroxylation was 4 times
higher than that of PFC10 (67/min/P450, Figure 3A).25 The
coupling efficiency of PFC9-L-Leu (36%) for propane
hydroxylation was increased 2 times in comparison with that
of PFC10 (18%),25 but it was lower than that of P450PMO
(94%).37 The higher affinity could be one of the possible
reasons for the higher rates of the hydroxylation reaction. It is
interesting to note that compounds with the L-amino acid were
better than those with the D enantiomer (Figure 3A and Table
S3, Supporting Information). PFC9 derivatives modified by L-
amino acids with larger side chains showed a tendency to give
higher PFRs (more than 200/min/P450), which suggested a
Data Collection and Refinement. Crystals were flash-
cooled by liquid nitrogen. X-ray diffraction data sets were
collected on the beamline BL26B2 equipped with a MAR225
CCD detector at SPring-8 (Hyogo, Japan) with a 1.0 Å
wavelength at 100 K. The program HKL200030 was used for
integration of diffraction intensities and scaling. The structure
was solved by molecular replacement with MolRep.31 The
structure of P450BM3 with N-palmitoylmethionine (1ZO9)
was used as a search model. Model building and refinement
were performed by using COOT32 and REFMAC5.33 The N-
perfluorononanoyl-L-tryptophan model was generated by using
153
dx.doi.org/10.1021/cs501592f | ACS Catal. 2015, 5, 150−156

ACS Catalysis
Research Article
significant steric effect of the side chain of the amino acid
(Figure 3A and Table S4, Supporting Information). Moreover,
a small amount of 1-propanol (23/min/P450) was detected
under these conditions (Tables S2−S5, Supporting Informa-
tion), whereas the hydroxylation of the unreactive primary sp3
C−H bond was observed with P450BM3 and PFCs only when
a high pressure of propane was applied (Figure S5, Supporting
Information)27 or when the reaction was performed by using
engineered P450BM3 and P450cam (Figure S8, Supporting
Information).38−40
The hydroxylation of ethane was then examined under
similar conditions by using NADPH. Because control experi-
ments using nitrogen gas instead of ethane showed a small
amount of ethanol formation due to a side reaction, we
calculated the formation rate of ethanol by ethane hydrox-
ylation (4−15/min/P450) after subtracting the background
ethanol (Figure S6 and Table S6, Supporting Information).
The rates were 6−22 times faster than that of the reaction
performed under high-pressure ethane (500 kPa) using PFC10
as the decoy molecule (0.67/min/P450)27 and were 10−37
times faster than that obtained with the P450BM3 variant
(P450PMO, 0.4/min/P450).39 It is not clear at this time how the
background ethanol was formed; however, decomposition of
NADPH could afford ethanol. Therefore, we also examined
ethane hydroxylation by employing a glucose dehydrogenase
promoted NADPH regeneration system to reduce the
concentration of NADPH in situ.41 The regeneration system
also promoted ethane hydroxylation, with a PFR range of 8−
45/min/P450 (Figure 3B and Figure S7 and Table S7,
Supporting Information). PFC9-L-Leu gave the largest PFR
(45/min/P450), and this activity was comparable with that
obtained with the best reported P450 variant (four mutations of
P450cam, 78.2/min/P450) for ethane hydroxylation (Figure
S8, Supporting Information).38 In addition, a long-term
reaction of ethane hydroxylation was carried out for 5 h in
the presence of PFC9-L-Leu together with an NADPH
Figure 4. (A) Crystal structure of P450BM3 with PFC9-L-Trp (red
sphere). (B) Binding modes of PFC9-L-Trp to P450BM3. An Fo − Fc
electron density map contoured at the 3.0 level omitting PFC9-L-Trp
is shown in light blue mesh. (C) The 2Fo − Fc electron-density map
around the active-site pocket. (D) Substrate-induced rotation of Arg47
in an increasing order of PAM- (palmitoleic acid, light blue), NPG-
(N-palmitoylglycine, magenta), NPM- (N-palmitoylmethionine, yel-
low), and PFC9-L-Trp-bound P450BM3 (green);. (E) Conformational
change of Phe87 in substrate-free P450BM3 (cyan), PFC9-L-Trp-
bound P450BM3 (pink), and NPG-bound P450BM3 (sky blue). (F)
DMSO-bound structure.
regeneration system to give a TON of 292 μmol/(μmol of
39
P450), which is lower than that observed by P450PMO
.
The
hydroxylation of cyclohexane was also examined by using the
second generation of decoy molecules, resulting in PFRs higher
than those obtained by using the first-generation decoy
molecules (Figure 3C).25
the bound forms of P450BM3 were also observed in the I, H,
and B′ helices and related loops (Figure S10B, Supporting
Information). More importantly, the quasi-open conformation
provides space that allows access to an additional small
substrate. Furthermore, a DMSO molecule unambiguously
occupied the sixth ligand position of the heme instead of a
water molecule, according to a 2Fo − Fc electron density map
contoured at 1σ around this region (Figure 4C), even though
only 1% DMSO was present (to improve the solubility of
PFC9-L-Trp). The distance between the sulfur atom of DMSO
and the heme iron is 2.4 Å (Figure S11A, Supporting
Information).43 This DMSO is also close to PFC9-L-Trp (4.2
Å), Phe87 (3.9 Å), Ala264 (3.2 Å), Thr268 (3.1 Å), and Ala328
(3.8 Å) (Figure S11B, Supporting Information). Considering
the similar size of DMSO and small alkanes, this result provides
evidence that the short chain length of PFC9-L-Trp allows
accommodation of the target substrates near the heme iron.
Interestingly, water molecules, which are usually observed near
the active center upon binding of long chain fatty acids (1JPZ
and 1ZO9), are not observed, possibly because of the increased
hydrophobic interaction of decoy molecules with the substrate-
binding channel upon the introduction of fluorine (Figure S12,
Supporting Information). The results are consistent with the
observation that the binding of decoy molecules leads to an
increased proportion of high-spin heme iron; thus, the
The high affinity of this series of N-perfluoroacyl amino acids
to P450BM3 (Table 1 and Figures S13 and S14, Supporting
Information) encouraged us to prepare cocrystals of P450BM3
and N-perfluoroacyl amino acids. As a result, we succeeded in
crystallizing a PFC9-L-Trp-bound form of P450BM3 and
obtained crystals that diffracted at 1.8 Å resolution (Figure
4A). In the fatty acid binding channel of P450BM3, clear
electron density that was readily assignable to PFC9-L-Trp was
observed, and the fluorine atoms were clearly distinguished
from proton atoms of fatty acids (Figure 4B and Figure S9,
Supporting Information). PFC9-L-Trp was bound to P450BM3
through hydrogen bond interactions with three amino acids:
Tyr51, Gln73, and Ala74 (Figure 4B). The overall structure has
a quasi-open conformation that is between the open
conformation (substrate-free, PDB code 2HPD)17 and the
closed substrate-bound conformations (PDB codes 1FAG
(palmitoleic acid),22 1JPZ (N-palmitoylglycine),18 and 1ZO9
(N-palmitoylmethionine)42). Clear changes in the quasi-open
conformation from the open conformation of 2HPD are found
in the F and G helices and the F/G loop (Figure S10A,
Supporting Information). Large differences in comparison with
154
dx.doi.org/10.1021/cs501592f | ACS Catal. 2015, 5, 150−156

ACS Catalysis
Research Article
reduction of ferric heme is expected to take place readily in the
presence of NADPH.44
modification of the amino acids to elongate the terminal of
PFCs is also expected to improve the binding affinity as well as
the solubility of decoy molecules.
Conformation changes of key residues in the PFC9-L-Trp-
bound P450BM3 provide clear evidence for the transition from
a resting state of P450BM3 (substrate free) to a “substrate-
bound-like” state upon binding of the decoy molecule. One
striking difference between the decoy molecule bound form and
the native substrate bound forms is found for Arg47, which is a
polar residue close to the entrance of the substrate access
channel. This residue moves toward the surface of the protein
and away from the substrate (Figure 4D), which may allow the
admission of a small substrate. Moreover, a hydrophobic side
chain of the decoy molecule occupies the space provided by the
rotation of Arg47. Therefore, the water network environment
in the entrance of the substrate-binding channel might be
ASSOCIATED CONTENT
* Supporting Information
The following file is available free of charge on the ACS
■
S
Publications website at DOI: 10.1021/cs501592f.
Materials, instruments, and experimental details, Figures
S1−S16, and Tables S1−S9 (PDF)
AUTHOR INFORMATION
Corresponding Authors
*E-mail for O.S.: shoji.osami@a.mbox.nagoya-u.ac.jp.
*E-mail for Y.W.: p47297a@nucc.cc.nagoya-u.ac.jp.
■
disturbed, which makes the entrance more hydrophobic.42
A
second significant change was observed in Phe87, a key residue
in the active site, which is 75° inclined to the heme plane,
whereas the DMSO binding may also contribute to this
orientation (Figures 4C). This perpendicular orientation is
similar to that observed in the substrate-free structure (2HPD)
but differs from the parallel orientation found in the N-
palmitoyl glycine-bound structure (1JPZ) (Figures 4E). The
perpendicular orientation of Phe87 reduces steric conflict with
the bound DMSO (Figure 4F). Because the structure of
DMSO-bound P450BM3 could be regarded as a model
structure of small alkane-bound P450BM3, the perpendicular
orientation of Phe87 would contribute to the accommodation
of small alkanes between Phe87 and Ala264. The crystal
structure analysis also reveals why the medium chain length of
the decoy molecule was effective, while decoy molecules having
longer alkyl chains showed better binding affinity to P450BM3.
The crystal structure of the PFC9-L-Trp-bound form of
P450BM3 indicated that the terminal perfluoromethyl group
of PFC9-L-Trp was close to DMSO (4.2 Å, Figure S11B,
Supporting Information) coordinated to the heme iron. This
observation clearly indicates that, in the case of decoy
molecules with alkyl chains longer than nine carbon atoms
such as PFC10-L-Leu and PFC11-L-Leu, the binding site for
gaseous alkane would be occupied by the alkyl chain of decoy
molecules or the access of gaseous alkane would be disturbed
by the alkyl chain of decoy molecules, both of which would lead
to lower catalytic activities and higher uncoupling reactions
(Table S5, Supporting Information),45 while the binding
affinities of PFC10-L-Leu and PFC11-L-Leu are higher than
that of PFC9-L-Leu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by a Grant-in-Aid for Scientific
Research (S) to Y.W. (24225004), Grants-in-Aid for Scientific
Research on Innovative Areas “Molecular Activation Directed
toward Straightforward Synthesis” to Y.S. (22105012) and to
O.S. (25105724), and a Grant-in-Aid for Young Scientists (A)
to O.S. (21685018) from the Ministry of Education, Culture,
Sports, Science, and Technology (MEXT) of Japan. We thank
Dr. Go Ueno, Dr. Yuki Nakamura, and Dr. Hironori Murakami
for their assistance with the data collection at SPring-8.
REFERENCES
■
(1) Rostrup-Nielsen, J. R. In Sustainable Strategies for the upgrading of
natural gas: fundamentals, challenges, and opportunities; Derouane, E.,
Parmon, V., Lemos, F., Ribeiro, F. R., Eds.; Springer: Dordrecth, The
Netherlands, 2005.
(2) Hashiguchi, B. G.; Konnick, M. M.; Bischof, S. M.; Gustafson, S.
J.; Devarajanm, D.; Gunsalus, N.; Ess, D. H.; Periana, R. A. Science
2014, 343, 1232−1237.
(3) Labinger, J. A.; Bercaw, J. E. Nature 2002, 417, 507−514.
(4) Luo, Y. R. Handbook of bond dissociation energies in organic
compounds; CRC Press: New York, 2003.
(5) Balasubramanian, R.; Smith, S. M.; Rawat, S.; Yatsunyk, L. A.
Nature 2010, 465, 115−119.
(6) Tinberg, C. E.; Lippard, S. J. Acc. Chem. Res. 2011, 44, 280−288.
(7) Nelson, D. R. Hum. Genomics 2009, 4, 59−65.
(8) Ortiz de Montellano, P. R. Cytochrome P450: structure,
mechanism, and biochemistry, 3rd ed.; Kluwer Academic/Plenum
Publishers: New York, 2005.
(9) Ortiz de Montellano, P. R. Chem. Rev. 2009, 110, 932−948.
(10) Petrik, I. D.; Liu, J.; Lu, Y. Curr. Opin. Chem. Biol. 2014, 19, 67−
75.
CONCLUSION
■
In conclusion, we have demonstrated that N-perfluoroacyl
amino acids strongly activate wild-type P450BM3 for the
hydroxylation of inert alkanes. PFC9-L-Leu gave PFRs for the
hydroxylation of secondary and primary C−H bonds of alkanes
that were comparable with those of the best P450 variants
prepared by multiple-round mutagenesis. The crystal structure
of the PFC9-L-Trp-bound form of P450BM3 revealed the active
site re-formation and provided mechanistic insight into the
activation of wild-type enzyme by decoy molecules. We
conclude that the catalytic activity for gaseous alkane
hydroxylation by wild-type P450BM3 would be improved
further by optimizing the structure of the decoy molecule.
Although we have focused herein on natural amino acids for the
modification of the carboxylate group of PFCs, non-natural
amino acids may further improve the catalytic activity. Further
(11) Ueno, T., Watanabe, Y. Coordination chemistry in protein cages:
principles, design and applications; Wiley: Hoboken, NJ, 2013.
(12) Narayan, A. R. H.; Sherman, D. H. Science 2013, 339, 283−284.
(13) Caswell, J. M.; O’Neill, M.; Taylor, S. J.; Moody, T. S. Curr.
Opin. Chem. Biol. 2013, 17, 271−275.
(14) Coelho, P. S.; Brustad, E. M.; Kannan, A.; Arnold, F. H. Science
2013, 339, 307−310.
(15) Kille, S.; Zilly, F. E.; Acevedo, J. P.; Reetz, M. T. Nat. Chem.
2011, 3, 738−743.
(16) Zhang, K.; Shafer, B. M.; Demars, M. D., II; Stern, H. A.; Fasan,
R. J. Am. Chem. Soc. 2012, 134, 18695−18704.
(17) Ravichandran, K. G.; Boddupalli, S. S.; Hasemann, C. A.;
Peterson, J. A.; Deisenhofer, J. Science 1993, 261, 731−736.
(18) Haines, D. C.; Tomchick, D. R.; Machius, M.; Peterson, J. A.
Biochemistry 2001, 40, 13456−13465.
155
dx.doi.org/10.1021/cs501592f | ACS Catal. 2015, 5, 150−156

ACS Catalysis
Research Article
(19) Rittle, J.; Green, M. T. Science 2010, 330, 933−937.
(20) Schlichting, I.; Berendzen, J.; Chu, K.; Stock, A. M.; Maves, S.
A.; Benson, D. E.; Sweet, R. M.; Ringe, D.; Petsko, G. A.; Sligar, S. G.
Science 2000, 287, 1615−1622.
subsequent binding of an oxygen molecule to initiate the formation of
compound I (Figure S16, Supporting Information).
(45) Kuper, J.; Wong, T. S.; Roccatano, D.; Wilmanns, M.;
Schwaneberg, U. J. Am. Chem. Soc. 2007, 129, 5786−5787.
(21) Whitehouse, C. J. C.; Bell, S. G.; Wong, L. L. Chem. Soc. Rev.
2012, 41, 1218−1260.
(22) Li, H.; Poulos, T. L. Nat. Struct. Biol. 1997, 4, 140−146.
(23) Boddupalli, S. S.; Estabrook, R. W.; Peterson, J. A. J. Biol. Chem.
1990, 265, 4233−4239.
(24) Miura, Y.; Fulco, A. Biochim. Biophys. Acta 1975, 388, 305−317.
(25) Kawakami, N.; Shoji, O.; Watanabe, Y. Angew. Chem., Int. Ed.
2011, 50, 5315−5318.
(26) Zilly, F. E.; Acevedo, J. P.; Augustyniak, W.; Deege, A.; Hausig,
U. W.; Reetz, M. T. Angew. Chem., Int. Ed. 2011, 50, 2720−2724;
Angew. Chem., Int. Ed. 2013, 52, 13503 (corrigendum about the system
being unavailable for methane oxidation but available for propane
oxidation) .
(27) Kawakami, N.; Shoji, O.; Watanabe, Y. Chem. Sci. 2013, 4,
2344−2348.
(28) Shoji, O.; Kunimatsu, T.; Kawakami, N.; Watanabe, Y. Angew.
Chem., Int. Ed. 2013, 52, 6606−6610.
(29) Carmichael, A. B.; Wong, L. L. Eur. J. Biochem. 2001, 268,
3117−3125.
(30) Otwinowski, Z.; Borek, D.; Majewski, W.; Minor, W. Acta.
Crystallogr. A 2003, 59, 228−234.
(31) Vagin, A.; Teplyakov, A. J. Appl. Crystallogr. 1997, 30, 1022−
1025.
(32) Emsley, P.; Cowtan, K. Acta Crystallogr., Sect. D 2004, 60,
2126−2132.
(33) Murshudov, G. N.; Skubak, P.; Lebedev, A. A.; Pannu, N. S.;
Steiner, R. A.; Nicholls, R. A.; Winn, M. D.; Long, F.; Vagin, A. A. Acta
Crystallogr., Sect. D 2011, 67, 355−367.
(34) Schttelkopf, A. W.; van Aalten, D. M. Acta Crystallogr. Sect. D
2004, 60, 1355−1363.
(35) DeLano, W. L. The Pymol Molecular Graphics System; http://
www.pymol.org.
(36) Watanabe, H., Fushihoshi, H., Ishida, S. (Fujifilm Co., JPN).
Preparation of perfluorocarboxylic derivatives. JP5-279307, October
26, 1993.
(37) Fasan, R.; Chen, M. M.; Crook, N. C.; Arnold, F. H. Angew.
Chem., Int. Ed. 2007, 46, 8414−8418.
(38) Xu, F.; Bell, S. G.; Lednik, J.; Insley, A.; Rao, Z.; Wong, L. L.
Angew. Chem., Int. Ed. 2005, 44, 4029−4032.
(39) Fasan, R.; Meharenna, Y. T.; Snow, C. D.; Poulos, T. L.; Arnold,
F. H. J. Mol. Biol. 2008, 383, 1069−1080.
(40) Chen, M. M.; Coelho, P. S.; Arnold, F. H. Adv. Synth. Catal.
2012, 354, 964−968.
(41) Weckbecker, A., Hummel, W. Microbial Enzymes and
Biotransformations; Humana Press: New York, 2005; Vol. 17, pp
225−238.
(42) Hegde, A.; Haines, D. C.; Bondlela, M.; Chen, B.; Schaffer, N.;
Tomchick, D. R.; Machius, M.; Nguyen, H.; Chowdhary, P. K.;
Stewart, L.; Lopez, C.; Peterson, J. A. Biochemistry 2007, 46, 14010−
14017.
(43) Matteis, F. D.; Ballou, D. P.; Coon, M. J.; Estabrook, R. W.;
Haines, D. C. Biochem. Pharmacol. 2012, 84, 374−382.
(44) Spectral titrations indicate that N-perfluoroacyl amino acids
bind to P450 with higher affinity in comparison to that of PFCs, as
evidenced by the estimated dissociation constants (Table 1 and
Figures S13 and S14, Supporting Information).25 For example, a small
amount of the heme iron in the high-spin state is observed upon
addition of decoy molecules (ca. 7% for PFC9-gly), and the ratio of
the high-spin heme iron increases to ca. 12% in the presence of
cyclohexane (Figure S15, Supporting Information). A ferrous CO
complex is formed when NADPH is added to a CO-saturated buffer
solution containing P450BM3 and N-perfluoroacyl amino acids. These
results suggest that tight binding of decoy molecules induces the
change of the spin state, followed by the reduction of ferric heme and
156
dx.doi.org/10.1021/cs501592f | ACS Catal. 2015, 5, 150−156